Reliability of measuring first and second metatarsal and toe length

1. Introduction 

Measurement of the length of the first metatarsal relative to the second metatarsal has been used in a number of studies to examine the relationship between length differences and the occurrence of foot problems [1], [2], [3], [4], [5], [6], [7]. It has been shown that a shorter first metatarsal transfers extra load onto the second metatarsal during walking [7] and this mechanism has been hypothesized as the potential cause for increased incidence of metatarsal stress fractures [8], increased foot pain, hallux rigidus [5], tarsometatarsal arthrosis [1], and hallux mobility [3].

In the population of elite classical dancers the relative length of the first metatarsal to the second metatarsal is regularly screened and recorded [9], [10], [11]. The dancer with a shorter first metatarsal is monitored to ensure that they maintain a satisfactory line and concentrate on technique in an attempt to avoid more proximal injuries [12].

Despite the interest in the difference between metatarsal and toe lengths, there is limited evidence that describes reliable, clinically applicable techniques of such foot measurements. Several studies have utilized an X-ray method to determine the actual and relative lengths of the metatarsals [1], [2], [4], [13]. However, this form of measurement is not an appropriate screening tool for clinical use due to cost, ionizing radiation exposure and accessibility. A more clinically appropriate method used in one study was a circumferential sketch of the foot obtained in weight-bearing, comparing the ends of the first two toes [5]. Unfortunately, this method was not well described, there was no indication of reliability, and the validity of measuring ends of the toes as an indicator of metatarsal length must also be questioned. Rodgers and Cavanagh [7] used palpation of the first and second metatarsal heads and then marked the heads with adhesive paper circles before taking a forensic footprint during walking. No reliability analysis was performed on this technique. Palpation of the metatarsals in an effort to determine the position of the metatarsal heads (or metatarsal formula) has been shown to be valid when compared with X-ray examination of metatarsal formula [14]. The authors reported that no statistical difference was found between the relative metatarsal position established using palpation and on X-ray, however the palpation techniques did not accurately determine relative metatarsal position in all cases.

Only one clinical method has been described in the literature to assess the relative lengths of the first and second metatarsals [3]. This method uses a calliper to determine the relative lengths of the first two metatarsals, measuring from the navicular tubercle to each metatarsal head. This technique provides details of the length of each metatarsal relative to the starting point at the navicular tubercle. Although simple to administer, the method was found to have poor inter-rater reliability (ICC=0.36). Intra-rater reliability was not reported.

Considering the dearth of reliable measurement techniques available and the ongoing research and clinical interest in the relative lengths of the first and second metatarsals, the aims of this research were to develop a method of measuring the difference between the ends of the first and second toes and a method of measuring the difference between the end of the metatarsals and test the intra-rater reliability of these methods. A further aim was to test the intra-rater reliability of the method described by Glasoe et al. [3] for measuring the difference between the relative lengths of the first and second metatarsal.

2. Methods 

2.2. Standardised markings and foot position 

For the three techniques examined, standardised foot markings and a standardised foot position was attained prior to testing. The medial aspect of the navicular tubercle and dorsal crease of the first and second metatarsophalangeal joints were palpated and marks made at each of these bony landmarks. A line was drawn down the middle of second toenail on each foot and a vertical line was drawn to bisect the posterior calcaneum. The centre of the calcaneum was determined using two vertical points made with a sliding vernier calliper. The first at the base of the calcaneum, the second 5cm above this point, a line was then drawn between these two points.

The participant then stood on a piece of graph paper (297mm×420mm) that was placed on a piece of plywood to standardise the surface. Participants placed their heel against the edge of a ruler, which was lined up with the single horizontal line on the graph paper (Fig. 1a). The foot was then positioned so that the marks on the calcaneum and second toe were in a straight line on the graph paper (Fig. 1b and c). This allowed the foot to be in a parallel position with the toes facing straight forward. Toes that were clawed or crooked were repositioned by the examiner to ensure they were straight.

View Large Image Download to PowerPoint

Fig. 1. Standardising foot position.

2.3. Technique 1: measurement of difference between ends of the first two toes 

With the foot in the standardised position, a carpenter's square was placed up against the end of the first toe and aligned with a line of the graph paper (Fig. 2a). This was to ensure that the line was straight and at the very end of the toe. The edge of a paint spatula tool was then stamped on a black ink stamp pad (Fig. 2b) and slid down between the toe and the edge of the square, making a black line parallel to the graph paper lines at the end of the toe (Fig. 2c). This was then repeated at the end of the second toe. The difference between the ends of the first and second toes was determined by measuring the distance between these two black lines and recorded in millimetres (Fig. 2d).

View Large Image Download to PowerPoint

Fig. 2. Technique one. (a) Carpenter's square against the toe, (b) INK on paint scraper, (c) marking the end of the toe and (d) measuring the distance between ends of first two toes.

2.4. Technique 2: measurement of difference between the ends of the metatarsals 

Keeping the foot in the standardised position, a red line was ruled horizontally across the page 2cm away from the initial mark made at the end of the first toe. A sliding vernier calliper was used to measure the distance in millimetres between the red line and the mark made at the dorsal crease of the metatarsophalangeal joint (Fig. 3). This was then repeated opposite the second toe and a measurement taken from the mark made at the end of the second metatarsal. The distance of the first metatarsal head from the external mark was subtracted from the distance of the second metatarsal head from the mark to provide the difference between the lengths of the two metatarsal.

View Large Image Download to PowerPoint

Fig. 3. Measuring the distance to metatarsal heads.

3. Results 

The means and standard deviations of the difference between first and second metatarsal and toe length using each measurement technique can be seen in Table 1. Coefficients of reliability demonstrated that technique one proved to be highly reliable, technique two was moderately reliable and technique three showed good reliability according to the ICCs [17] (Table 1). Portney and Watkins [17] suggest that ICC values over 0.75 are indicative of good reliability and values below 0.75 indicate poor to moderate reliability.

Table 1. Retest reliability of measurement of first toe and metatarsal length relative to the second toe and metatarsal length

	Technique	Mean test score±S.D.	Mean retest score±S.D.	ICC2,1	Mean difference	S.D. difference	95% CI (group mean)	95% CI (individual)
	Technique 1	4.39±4.60	4.29±4.32	0.98	0.10	0.87	−0.19 to 0.39	−1.66 to 1.86
	Technique 2	−3.25±3.16	−3.18±3.73	0.67	−0.07	2.84	−1.03 to 0.89	−5.81 to 5.67
	Technique 3	9.32±3.17	9.56±3.42	0.76	−0.24	2.31	−1.02 to 0.54	−4.91 to 4.43

Note: Mean difference equals test minus retest scores. Toe and metatarsal length differences are in millimetres.

The mean group 95% CI for each technique crossed zero, indicating that there was no significant difference between the test and retest means and no systematic error occurred between the testing sessions [15], [18] (Table 1).

All three techniques displayed narrow 95% CI for the group mean score. For technique one the individual score 95% CI limits were also narrow, suggesting an acceptable degree of agreement between testing sessions and a relatively precise measurement [16]. For technique two and three the individual 95% CI limits were wider, indicating that the measurement of first and second metatarsal length differences for an individual could vary by up to 12mm (technique two) or 10mm (technique three) between a testing session. These wide intervals indicate measurement error associated with the techniques.

4. Discussion 

The purpose of this study was to determine intra-rater reliability of three different techniques for measuring anthropometric data of the feet. The findings indicate excellent reliability (ICC=0.98) when using technique one to measure the difference between the ends of the toes. In addition, the small range of the group mean and individual score 95% CI indicate an accurate method of measurement that would be sufficiently reliable for use in individual screening by the clinician and for use in evaluating groups of subjects.

Good reliability was indicated in the results for technique three (ICC=0.76). This technique measured the difference between the ends of the first and second metatarsals taken from the navicular tubercle, a measurement technique described by Glasoe et al. [3]. For technique two, measuring the difference between the ends of the first and second metatarsals from a mark outside of the foot, moderate reliability was found (ICC=0.67).

Techniques two and three both displayed wider 95% CI for the individual scores and might not be considered sufficiently reliable for use in the clinical setting. In particular, technique two showed large individual score 95% CI limits relative to test scores (test one mean score=−3.25mm). This lack of precision would be clinically unacceptable considering the small difference in metatarsal length observed at each test. Technique three yields greater differences in metatarsal length (test one mean score=9.32mm) and therefore despite the wide individual score CI limits, showed greater measurement precision than technique two.

The narrow group mean 95% CI limits for technique one and three, relative to the test scores, indicates that these measures would be reliable for the evaluation of metatarsal and toe length in groups of subjects. Such techniques could be useful for evaluating the influence of toe and metatarsal length on the incidence of foot disorders such as stress fractures or hallux valgus.

The reduced reliability of technique two and three could be explained by examiner inconsistencies and measurement error. Repeated and accurate marking of the ends of the metatarsal heads and navicular tubercle in technique three is difficult due to the rounded shape of the bony prominences [3]. The nature of skin moving over the bones could also represent another source of error, as the position on the skin where the bony landmark is felt, may move as the finger is removed and a mark is made. Another source of error is the amount of weight being placed on the foot being measured. The participant was asked to put most of their weight through this foot but they needed to be able to balance, so weight was also being distributed through their other foot. As weight increases through the foot, the bones move to absorb and distribute the greater weight. This variability of weight placement could affect the measurements between tests. The use of a scale for the participant to place their other foot on could standardise the amount of weight put through the foot to be measured.

The increased number of elements used in technique two could explain the inflated measurement error. This technique relied on accurate palpation and marking of the metatarsal heads and accurate placement of the carpenter's square along the ruled line. Placement of the jaws of the calliper on the metatarsal mark and against the set-square could also present opportunities for error. The angle the calliper is placed at can affect the actual distance read and the shape of the calliper made it difficult to have it at the same angle for each measurement. All of these opportunities for error have contributed to this being considered a technique that is too difficult to achieve good reliability even with a single tester.

Technique one and two were designed specifically for this study and technique three has been used in one other study [3]. These authors found this method of measurement to be unreliable between testers and commented on the difficulty in palpating the metatarsophalangeal joint line. Using a single tester appears to be a more reliable way of utilising this method, possibly due to the tester using the same palpatory techniques for finding the metatarsal head and navicular.

One of the major strengths of this study was that intra-rater reliability has been established for three relatively simple and inexpensive techniques for collecting anthropometric data. A reliable method has been devised to measure the difference between the ends of the first and second toes. Measuring the difference between the ends of the first two metatarsals is more difficult, but there are now two methods that have been described and further research may improve on these techniques. The method initially described by Glasoe et al. [3] as having poor inter-rater reliability can now be considered with a certain amount of confidence when using a single rater for measurement and particularly for use in groups of subjects.

One of the major limitations of this study is its inability to determine the validity of each of the techniques used. This could be addressed with the use of X-rays or cadaver specimens, to determine if the parameters used to obtain the measurements were indeed finding the differences between the ends of the metatarsals and the toes. Another consideration is that the results may not be generalisable to other populations of subjects, such as those with toe deformities, and other populations of testers, such as those with limited musculoskeletal assessment experience. The inter-rater reliability of these measures requires further investigation.

5. Conclusion 

A new technique devised by the authors for measuring the difference between the ends of the first two toes and a technique previously described for measuring the relative difference between the ends of the first two metatarsals [3] have been found to have good to excellent intra-rater reliability. Both methods are clinically relevant, inexpensive and non-invasive and could be used for collecting anthropometric data of the feet in future research or within the clinical setting during screening sessions.

Conflict of interest 

There are no known conflicts of interest.